<?xml version="1.0" encoding="UTF-8"?><xml><records><record><source-app name="Biblio" version="6.x">Drupal-Biblio</source-app><ref-type>10</ref-type><contributors><authors><author><style face="normal" font="default" size="100%">Pierre Enel</style></author><author><style face="normal" font="default" size="100%">Mehdi Khamassi</style></author><author><style face="normal" font="default" size="100%">Emmanuel Procyk</style></author><author><style face="normal" font="default" size="100%">Peter F. Dominey</style></author></authors></contributors><titles><title><style face="normal" font="default" size="100%">REINFORCEMENT LEARNING MODEL IN PROBABILISTICALLY REWARDED TASK</style></title><secondary-title><style face="normal" font="default" size="100%">Neurocomp</style></secondary-title></titles><keywords><keyword><style  face="normal" font="default" size="100%">anterior cingulate cortex</style></keyword><keyword><style  face="normal" font="default" size="100%">computational model</style></keyword><keyword><style  face="normal" font="default" size="100%">decision making</style></keyword><keyword><style  face="normal" font="default" size="100%">exploration</style></keyword><keyword><style  face="normal" font="default" size="100%">neural simulation</style></keyword><keyword><style  face="normal" font="default" size="100%">prefrontal cortex</style></keyword></keywords><dates><year><style  face="normal" font="default" size="100%">2010</style></year></dates><urls><web-urls><url><style face="normal" font="default" size="100%">http://hal.archives-ouvertes.fr/hal-00553440/en/</style></url></web-urls><related-urls><url><style face="normal" font="default" size="100%">http://organic.elis.ugent.be/sites/reservoir-computing.org/files/sites/organic.elis.ugent.be/files/Enel2010_NEUROCOMP2010.pdf</style></url></related-urls></urls><pub-location><style face="normal" font="default" size="100%">Lyon, France</style></pub-location><abstract><style face="normal" font="default" size="100%">&lt;p&gt;Adapting resource seeking behavior is of&lt;br /&gt;
primary importance in survival. Then, balancing&lt;br /&gt;
exploration and exploitation of discovered resources is&lt;br /&gt;
at the core of adaptation to the environment. The&lt;br /&gt;
reinforcement learning theoretical framework has been&lt;br /&gt;
elaborated to formalize such reward seeking behavior.&lt;br /&gt;
Biologically plausible models based on this algorithm&lt;br /&gt;
have flourished recently. Among them, a neural&lt;br /&gt;
network model was developed to investigate the&lt;br /&gt;
functions of the anterior cingulate cortex (ACC) and the&lt;br /&gt;
dorsolateral prefrontal cortex (DLPFC) involved in&lt;br /&gt;
action valuation and action selection, respectively [1].&lt;br /&gt;
This model proposes a method to regulate dynamically&lt;br /&gt;
the exploration inspired by literature on meta-learning&lt;br /&gt;
in order to solve dynamically the exploration/&lt;br /&gt;
exploitation trade-off [2]. This model performed well in&lt;br /&gt;
a deterministic problem solving task (PST). Our goal&lt;br /&gt;
was to demonstrate that the model is generalizable to a&lt;br /&gt;
more ecological PST with probabilistically dispensed&lt;br /&gt;
rewards. The model was tested with its preset learning&lt;br /&gt;
rate / exploration rate / initial action values and then&lt;br /&gt;
optimized with search of the parameters space. The&lt;br /&gt;
initial values of model's parameters proved to be good&lt;br /&gt;
however not optimal for the new task. Interestingly, the&lt;br /&gt;
model's performance is very dependent on the initial&lt;br /&gt;
action values.&lt;/p&gt;</style></abstract></record></records></xml>